AI Foundations aims to enhance player experiences through ML and AI, with the ML Bots team specifically developing Game Understanding Agents to power in-game player experiences across Riot's titles by creating autonomous agents that can play, understand, and adapt like real players.
Requirements
- Extensive experience (8+ years) delivering ML systems in production, including reinforcement learning, imitation learning, or simulation-based training in rich, interactive environments such as game worlds or multi-agent simulations.
- Proven ability to design modeling strategies and architectures adopted across multiple games or interactive products.
- Expertise in developing predictive features and signals from gameplay telemetry, simulation data, or other complex interactive environments.
- Strong track record building and optimizing agent-based systems or world models for dynamic, player-facing environments.
- Hands-on experience with relevant ML methods including reinforcement learning and imitation learning (such as behavior cloning and inverse reinforcement learning), on-/off-policy algorithms, policy gradient methods, behavior shaping, and hybrid systems that combine learned policies with rule-based or scripted components.
- Track record of incorporating human considerations into AI applications, such as responsible AI practices and human-computer interaction or UX best practices.
- Experience mentoring engineers and collaborating with cross-disciplinary teams
Responsibilities
- Lead the modeling strategy for ML Bots across multiple games, focusing on training agents that can understand game state, make decisions, and act in ways that create compelling player experiences.
- Develop predictive features and signals from gameplay telemetry, unstructured game data, and simulation outputs, ensuring quality, interpretability, and reliability.
- Design and implement ML systems using methods including reinforcement learning and imitation learning (e.g., behavior cloning, inverse reinforcement learning), on-/off-policy algorithms, policy gradient methods, behavior shaping, and hybrid systems that combine learned policies with rule-based or scripted components.
- Define evaluation frameworks for game AI that balance generalizable approaches with genre-specific metrics, adapting methods for the needs of each title.
- Ensure the safety, fairness, and trustworthiness of autonomous agents operating in live player environments.
- Mentor senior and staff-level ML engineers in advanced ML for game AI and architectural decision-making.
- Collaborate with game and platform engineers, along with UX teams, to integrate models into production systems in ways that enhance player experience and maintain operational reliability.
Other
- player empathy and care about players' experiences
- collaborative spirit
- decision-making that prioritizes the delight of players
- Safeguarding confidential and sensitive Company data
- Communication with others, including Rioters and third parties such as vendors, and/or players, including minors
